KR20220033165A - System for monitoring image employoing sensing fire with filter - Google Patents

Abstract

An embodiment relates to a fire detection video monitoring system with a filter. Specifically, the video monitoring system includes: a camera for capturing an image of a pre-specified target; and an information processing device for determining whether a fire has occurred on the basis of a capturing result obtained by the camera. Accordingly, when a situation similar to a fire occurs, a judgment is first made through video analysis, and a fire alarm is issued by accurately determining whether there is a fire through a filter exchange performed secondly.

Description

Fire detection image monitoring system having a filter {System for monitoring image employoing sensing fire with filter}

The content disclosed in this specification relates to the technical field of a fire detection video surveillance system, and more particularly, to a system for determining whether a fire exists by using a fire detection image captured by a camera.

Unless otherwise indicated herein, the material described in this section is not prior art to the claims of this application, and inclusion in this section is not an admission that it is prior art.

In general, in order to prevent great damage to fires that may occur in shipbuilding industries, offshore plants, and petrochemical plants, a fire detector for detecting whether a fire has occurred is installed.

A smoke detector that detects smoke caused by a fire and a heat detector that detects heat (temperature) from a fire are used as fire detectors. is also being used.

However, the conventional fire detector is a method of sounding an alarm when a flame that has already spread widely and smoke is generated heavily. It has a problem that cannot be effectively prevented.

Accordingly, the shipbuilding industry urgently needs to develop equipment to secure the safety of LNG bunkering due to the increase in demand for LNG fuel. In other words, the construction of LNG bunkering infrastructure and securing safety in industrial sites handling explosive gas are very important issues.

In addition, in the related art, it is a monitoring system that detects gas and flame at the same time, and since it is a system including all of a smoke detector, a thermal camera, a CCTV, and an IR gas detector, many sensors are required. There was a problem with

The prior art documents of this background include the following patent documents.

(Patent Document 0001) KR10-1869442 B1

For reference, Patent Document 1 relates to a system for monitoring a fire using a camera monitoring system and an automatic smoke detection algorithm based on image processing technology.

The disclosed content is a fire with a filter that can provide a system that makes a judgment through image analysis first, and accurately determines whether a fire exists through secondary filter replacement when a situation similar to the occurrence of a fire occurs, and issues a fire alarm To provide a detection video surveillance system.

A fire detection video surveillance system having a filter according to an embodiment,

a camera for taking an image of a predetermined target; and

an information processing device for determining whether a fire has occurred based on a result captured by the camera; contains

And, the camera,

a) The first filter replacement unit, which is formed by sequentially forming the filter replacement driving unit (A), the transparent unit (B), and the infrared cut filter unit (C), the transparent unit (B), the infrared transmitting unit (D), and the filter replacement driving unit (A) ) is formed in this order, and the first filter exchange part and the second filter exchange part connect each transparent part (B) in the vertical direction to the opposite transparent part (B) and the infrared transmitting part (D). ) or by partially overlapping all of the infrared cut filter part (C),

b) In the case of the weekly monitoring mode, the first filter exchange unit is replaced with an infrared cut filter unit (C), and the second filter exchange unit is replaced with a transparent unit (B) to realize a color image,

c) In the case of night monitoring mode, the first filter exchange unit is replaced with a transparent part (B) and the second filter exchange part is replaced with a transparent part (B) to realize a black-and-white image,

d) In the case of fire analysis mode, the first filter exchange unit is a transparent part (B), and the second filter exchange part is filter exchanged with an infrared transmitting part (D) to realize a high-temperature analysis image,

The information processing device,

a′) when the light corresponding to the first light quantity preset in b) and c) is incident, instructing to replace the filter in d),

b') when the infrared region light incident by a') corresponds to a preset amount of light, determining a fire; is characterized by

According to embodiments, when a situation similar to the occurrence of a fire occurs, a system is provided for first determining through image analysis and accurately determining whether a fire exists through secondary filter replacement to issue a fire alarm.

Accordingly, there is a very excellent effect in analyzing the cause of a fire, such as an early warning of a fire, an ignition point of a fire, a direction of a fire, etc. through an initial diagnosis in case of a fire.

1 is a diagram for conceptually explaining a fire detection video surveillance system having a filter according to an embodiment;
2 and 3 are diagrams for explaining a fire detection video surveillance system according to an embodiment
4 is a flowchart sequentially illustrating the operation of the fire detection video surveillance system having a filter according to the embodiment of FIG. 3 ;

1 is a diagram for conceptually explaining a fire detection video surveillance system having a filter according to an embodiment.

As shown in Fig. 1, the fire detection video surveillance system having a filter according to an embodiment makes a judgment through image analysis first when a situation similar to the occurrence of a fire occurs, and determines whether a fire exists through secondary filter replacement It is a system that accurately determines and fires a fire alarm (a specific operation will be described with reference to FIGS. 2 and 3 below).

For example, the fire detection video monitoring system receives, stores and outputs video data from the camera 20 installed in the monitoring area such as a bank, a government office, a hotel, a factory, a parking lot, and a park through a communication network. It is a kind of control system that allows you to check the on-site situation. Such a fire monitoring system 30 may be installed in a security company, a police situation room, a fire prevention agency situation room, or the like.

Here, several cameras 20 may be installed in one monitoring area (eg, installed on each floor in a building), and one fire monitoring system 30 may provide image data from a plurality of cameras 20 installed in several monitoring areas. can be transmitted and monitored. However, for convenience of description, in FIG. 1 , a single camera 20 installed in one monitoring area and photographing a specific monitoring area is illustrated and described.

The camera 20 installed in the monitoring area acquires image data by photographing a specific monitoring area, converts the obtained image data into digital image data, and continuously transmits it to the fire monitoring system 30 through a communication network through a compression process do.

The fire monitoring system 30 receives the compressed image data from the camera 20 , processes it, and stores it in the storage medium 39 or outputs it to the display unit 38 so that the administrator can check it.

Here, a plurality of reference devices 10 are installed in the monitoring area photographed by the camera 20 . The reference device 10 refers to a device that is intentionally installed to enable image analysis by causing a physical/chemical change to change color or location when a fire occurs.

Additionally, a general aspect of such a fire monitoring system, that is, an information processing device for fire monitoring, will be described as follows.

Here, the camera 20 captures the corresponding monitoring area to obtain image data, digitally converts it, and adjusts the resolution to compress it. Compression of image data is performed by Wavelet, Joint Photographic coding experts group (JPEG), Moving picture experts group (MPEG-1, MPEG-2, MPEG-4, MPEG-7, MPEG-21...) , H264, fractal, etc., and compresses the image data to any one of 352×240, 720×240, 720×480, 800×600, 1024×768, 1280×1024 resolution, 30, Video data can be compressed at frame rates of 15, 10, or 5 fps.

When the image data receiving unit 31 receives the image data compressed and transmitted by the camera 20 in this way, the decoding unit 32 decompresses it and converts it into digital image data. The decoding unit 32 decompresses the compression corresponding to the compression method compressed for transmission by the camera 20 .

The D/A converter 33 (Digital to Analog converter) converts the digital image data decompressed by the decoding unit 32 into analog image data that can be output on a screen.

The storage unit 34 is provided to store the image data received by the image data receiving unit 31 in the storage medium 39 . The storage unit 34 preferably stores the compressed image data received through the image data receiving unit 31 in the storage medium 39, and matches the unique ID of the camera 20 that has transmitted the image data. By storing it, it is possible to check the image data received through which camera 20 in the future.

The storage medium 39 may be a conventional storage device such as a hard disk for storing data. Alternatively, the storage medium 39 is provided with a database management system (DBMS) that operates an image data D/B for storing image data received from the camera 20, or interoperates with a data server equipped with a DBMS. can be

The screen output unit 35 processes the analog image data converted by the D/A conversion unit 33 to be output through the display unit 38 (or the monitor of the manager terminal 40 ). Here, the fire monitoring system 30 may receive image data from a plurality of cameras 20 . For this purpose, the screen output unit 35 controls the arrangement or size of a screen to be output to the display unit 38 to process the output. do. For example, when image data is received from 16 cameras 20 , the screen output unit 35 determines the division and output positions and sizes of 16 screens to output 16 image data through the display unit 38 .

On the other hand, on the screen output from the display unit 38 through the screen output unit 35, not only the image of the surveillance area acquired by the camera 20, but also a text string indicating the position at which the image was acquired or the uniqueness of the camera 20 It may further include an ID, and may further include a warning message informing that a specific situation has occurred.

In addition, the position where the camera 20 is installed on the screen output unit 35 may be output in a predetermined area of the display unit 38 or through a separate display means using a matrix and a map. In addition, when a fire occurs, an icon displayed on the position of the corresponding camera 20 output using a matrix and a map may be flashed to notify the location of the situation to the manager.

Here, the display unit 38 may be an individual display unit or a unit on which a plurality of small monitors are loaded to form a large situation board, or may be a monitor associated with the manager terminal 40 according to implementation.

The fire determination unit 36 analyzes the analog image data converted by the D/A conversion unit 33 to determine whether a fire has occurred. That is, it is determined whether a fire has occurred by checking whether a change occurs in the reference device 10 due to the occurrence of a fire.

2 and 3 are diagrams for explaining a fire detection video surveillance system according to an embodiment.

Specifically, FIG. 2 is a block diagram illustrating the configuration of a fire detection video surveillance system having a filter according to an embodiment. And, FIG. 3 is a view for explaining a filter function applied to a fire detection video surveillance system having a filter according to an embodiment.

First, as shown in FIG. 2 , the fire detection video surveillance system of an embodiment includes, for example, a camera 100 that captures an image of a predetermined target, and a fire based on a result captured by the camera 100 . It includes an information processing device 200 for determining whether or not.

Here, as a specific example, the camera 100 includes a lens unit 101 , a primary image analysis unit 102 , a filter exchange unit 103 , a first transmission unit 104 , and a first receiving unit 105 . include

In addition, the information processing apparatus 200 includes a second receiving unit 201 , a secondary image analyzing unit 202 , a second transmitting unit 203 , and an alarm generating unit 204 .

In this state, a filter function according to an embodiment is performed (refer to FIG. 3).

That is, as shown in FIG. 3, the filter function according to an embodiment is installed by overlapping the first filter exchange unit and the second filter exchange unit first.

That is, the first filter replacement part, which is formed in order of the filter exchange driving part (A), the transparent part (B), and the infrared cut filter part (C), the transparent part (B), the infrared transmission part (D), and the filter exchange driving part (A) ) is formed in this order, and the first filter exchange part and the second filter exchange part connect each transparent part (B) in the vertical direction to the opposite transparent part (B) and the infrared transmitting part (D). ), or partially overlapping all of the infrared cut-off filter unit (C).

In this state, in the case of use as a normal daytime monitoring camera and fire monitoring, the first filter replacement unit is C, and the second filter replacement unit is B. So, through this, the daytime monitoring mode blocks infrared rays and receives visible light to realize color images (daytime monitoring mode: visible light transmission (wavelength: about 400nm ~ 720nm)).

And, in the case of use as a normal night surveillance camera and fire monitoring, unlike this, the first filter replacement part is B, and the second filter exchange part is B. So, through this, infrared and visible light are received in the night monitoring mode to realize black and white images (night monitoring mode: visible light + near-infrared transmission (wavelength: about 400nm ~ 1.2μm)).

Then, when strong light enters in these two cases, the first filter exchange unit 103 filters the first filter exchange unit B and the second filter exchange unit changes the filter D, and the image and data in this case are converted into an information processing device (200). Therefore, in the fire analysis mode, visible light is blocked and infrared is received and an image capable of analyzing a high temperature region is transmitted to the information processing device 200 (fire analysis mode: infrared transmission (wavelength: liquid 0.72 μm ~ 1.2) μm)).

Then, the information processing apparatus 200 analyzes whether or not there is a fire in the secondary image analysis unit 202 through the second reception unit 203 . Specifically, the information processing device 200 blocks the visible ray and the image collected by receiving infrared rays blocks visible ray light and transmits only light in the high temperature region, so a certain amount of transmitted infrared ray light is transmitted. If it exceeds, it is judged as fire.

So, as a result of the determination, when a fire occurs, the fact is transmitted to the alarm generating unit 204 to provide a fire alarm, and the current filter state is maintained through the second transmitting unit 203 and the first receiving unit 105 . do.

On the other hand, if a fire does not occur as a result of the determination, it instructs to change the filter to the day monitoring mode or the night monitoring mode through the second transmitter 203 and the first receiver 105 .

4 is a flowchart sequentially illustrating the operation of the fire detection video surveillance system having a filter according to the embodiment of FIG. 3 (refer to FIGS. 2 and 3 ).

As shown in Fig. 4, the fire detection video surveillance system of an embodiment first determines whether a fire has occurred based on a camera that shoots an image of a pre-specified target as in the existing one, and a result captured by the camera. It is assumed that an information processing device is included.

In this state, according to an embodiment, in the video surveillance system, the camera has the following configuration.

That is, the camera is

a) The first filter replacement unit, which is formed by sequentially forming the filter replacement driving unit (A), the transparent unit (B), and the infrared cut filter unit (C), the transparent unit (B), the infrared transmitting unit (D), and the filter replacement driving unit (A) ) is formed in this order, and the first filter exchange part and the second filter exchange part connect each transparent part (B) in the vertical direction to the opposite transparent part (B) and the infrared transmitting part (D). ) or partially overlapping all of the infrared cut filter unit (C) (see FIG. 3 ).

b) In this state, when the fire mode is the daytime monitoring mode (S401), the first filter replacement unit is replaced with an infrared cut filter unit (C), and the second filter replacement unit is replaced with a transparent unit (B) to realize a color image. becomes (S402).

c) In the case of night monitoring mode (S401), the first filter exchange unit is replaced with a transparent part (B) and the second filter exchange part is replaced with a transparent part (B) to realize a black-and-white image (S403).

Then, when the light corresponding to the first amount of light preset in b) and c) is incident, that is, when a bright light of a certain amount or more is received, the information processing apparatus instructs the filter to be replaced in step d).

Then, when the camera is in the fire analysis mode in this way (S401), the first filter exchange unit is replaced with a transparent part (B), and the second filter exchange part is replaced with an infrared ray transmitting part (D), so that the high-temperature analysis image is displayed. is implemented (S404).

Therefore, the information processing apparatus determines that the incident infrared light corresponds to a preset amount of light according to this implementation, that is, when the infrared light is greater than or equal to a certain amount (S405). In this case, the fire determination is performed by decomposing the transmitted infrared light for each unit and comparing it with a predetermined amount to determine the fire area.

Accordingly, through this, an embodiment provides a system that, when a situation similar to the occurrence of a fire occurs, first makes a judgment through image analysis, and accurately determines whether a fire exists through secondary filter replacement, and then issues a fire alarm.

Accordingly, there is a very excellent effect in analyzing the cause of a fire, such as an early warning of a fire, an ignition point of a fire, a direction of a fire, etc. through an initial diagnosis in case of a fire.

As described above, an embodiment is a first filter replacement unit in which the filter replacement driving unit (A), the transparent unit (B), and the infrared cut filter unit (C) are sequentially formed as described above for a camera used for fire monitoring. and a second filter exchange part in which the transparent part (B), the infrared ray transmitting part (D), and the filter exchange driving part (A) are formed in this order. And, also in this case, the first filter exchange unit and the second filter exchange unit vertically connect each transparent part (B) to the opposite side's transparent part (B) and infrared transmission part (D) or infrared cut filter part (C) All are partially overlapped and installed.

So, in one embodiment, in the case of the day monitoring mode, the first filter exchange unit is replaced with a) an infrared cut filter unit (C), and the second filter exchange unit is replaced with a transparent unit (B) to realize a color image.

And, b) in the night monitoring mode, the first filter exchange unit is replaced with a transparent part (B) and the second filter exchange part is replaced with a transparent part (B) to realize a black-and-white image.

In addition, in particular c) in the case of the fire analysis mode, the first filter exchange unit is replaced with a transparent part (B) and the second filter exchange part is replaced with an infrared ray transmitting part (D), so that an image for high temperature analysis is realized.

In this case, the information processing apparatus instructs a') to replace the filter in c) when the light corresponding to the first light quantity preset in a) and b) is incident.

And, b') When the infrared region light incident by a') corresponds to a preset amount of light, it is determined as a fire.

Accordingly, through this, an embodiment provides a system that, when a situation similar to the occurrence of a fire occurs, first makes a judgment through image analysis, and accurately determines whether a fire exists through secondary filter replacement, and then issues a fire alarm.

Accordingly, there is a very excellent effect in analyzing the cause of a fire, such as an early warning of a fire, an ignition point of a fire, a direction of a fire, etc. through an initial diagnosis in case of a fire.

Additionally, the filter function according to the embodiment also performs the bad weather mode by the above-described filter replacement unit.

Specifically, the following operation is performed.

First, the information processing apparatus instructs to change the filter in step c), that is, in the night monitoring mode, when light lower than the preset second light amount is incident in b), that is, in the day monitoring mode.

Then, when the camera is in the first bad weather mode according to this instruction, the first filter exchange unit is replaced with a transparent part (B) and the second filter exchange part is replaced with a transparent part (B) to realize a black-and-white image.

Therefore, the bad weather mode is performed through these black-and-white images.

In this case, the black-and-white image is not affected at all by rain, snow, sea fog, hail, etc. because infrared wavelengths pass through rain, snow, sea fog, hail, etc., which are bad weather, and thus a distinctive image is obtained.

In this regard, the filter function according to the embodiment has a form different from that described above.

The specific configuration is as follows.

First, in b), that is, in the day monitoring mode, the information processing device compares the sharpness value and the saturation value of the current color image with the sharpness value and the saturation value of the last previous color image of the current color image. In the above c), that is, in the night monitoring mode, the filter replacement is instructed.

Then, when the camera is in the second bad weather mode according to the instruction, the first filter exchange unit is replaced with a transparent part (B) and the second filter exchange part is replaced with a transparent part (B) to realize a black-and-white image.

On the other hand, in relation to the filter function of this embodiment, in the embodiment, it is preferable that the first filter exchange unit and the second filter exchange unit are configured at the front end or rear end of the lens unit of the camera.

On the other hand, additionally, the above embodiments detect a fire after switching the filter. In this case, the occurrence of a fire in a tunnel having a higher risk of fire is smoothly and effectively detected, and the frequency of false detection of a fire can be further reduced.

This operation is performed, for example, as follows.

That is, the operation first switches the filter, receives the image signal output from the camera, acquires image frames at regular time intervals, uses a window mask to subsample the image, or performs filtering for noise removal.

Next, a threshold is applied to stably detect a region where high heat is generated in the infrared image. In this case, the threshold value is set in advance for each unit by decomposing the light in the infrared region.

Then, by applying the threshold, a pixel having a high brightness value is continuously detected in the same pixel for each light divided for each unit, and an area having a high brightness value is detected using the connectivity of the detected binary pixels. In addition, from the region detected as a region having a high brightness value in the current frame, the number of pixels indicating the area of the region, the center position, the brightness average and standard deviation of the region, and the characteristics of the rectangular information including the moment region are calculated.

In this case, the threshold Th1 for detecting a pixel having a high temperature in the infrared image is obtained through, for example, a brightness histogram of the infrared image.

For example, it is set to a value between 120 and 190 in an infrared image having a brightness value of 256 steps.

In addition, the threshold Th2 for filtering a small-sized area detected by noise or the like is affected by a sub-sampling size and a distance from a location where the camera is installed.

In this case, the distance from the camera is determined to be farther from the camera as the distance from the camera is set by setting the exit position of the tunnel in the two-dimensional image, obtaining the distance in two dimensions to an arbitrary area, and having a smaller value for this distance.

And, the threshold value for detecting pixels having a high temperature is inversely proportional to the distance from the camera by linearly interpolating the minimum value 0 and the maximum value of the distance in the 2D image region between 120 and 190, that is, the position of each region. set to a different value depending on

On the other hand, the farther the distance from the camera is, the smaller the object appears, so the threshold for filtering a small area detected by noise, etc. .

Accordingly, by comparing the size change of each area detected from the current frame with the area detected from the recent frame, if the size continuously changes, it is determined as a fire, and if it does not change, it is determined as not a fire.

Accordingly, in the above embodiments, a region that is a high temperature region continuously over several frames in an infrared image and has a change in size or shape due to a flame is detected as a flame caused by a fire for each light in the infrared region decomposed for each unit. Accordingly, the above embodiments smoothly and effectively detect the occurrence of a fire in a tunnel or the like with an infrared camera installed in a tunnel having a higher risk in case of a fire, and further reduce the frequency of false detection of a fire.

For reference, in the above embodiments, a binary image by a threshold is obtained using Equations 1, 2, 3, 4, and 5 below, and an area having a high brightness value is continuously detected in the same area by calculating an accumulated image. do.

이다,Equation 1 is

am,

Here, I(x, y, t) is an image frame at time t, Th1 is a threshold value for detecting a pixel having a high temperature in an infrared image, and BI(x, y, t) is a threshold value Th1 in an infrared image It is a binary image of a pixel having a higher brightness value.

이다,And, Equation 2 is

am,

Here, QueueBI is obtained by storing the most recent maxQ1 binary images in a First-In First-Out method, and mod is the remainder operator.

이다.Also, Equation 3 is

am.

Here, SI(x, y, t) is a region having a high temperature in a region continuously overlapping by calculating a sum from QueueBI and applying a threshold.

이다.And, Equation 4 is

am.

이다.Also, Equation 5 is

am.

On the other hand, the fire

By using Equations 6, 7, and 8 below, features are extracted from a region with a continuously high brightness value, and the region in which the shape and size of the region fluctuates due to the movement of the flame is obtained from the infrared camera installed in the tunnel. Determine the flame generation area from the infrared image.

이다.Equation 6 is

am.

Here, htBlob(i, t) is a feature detected as a region having a consistently high temperature at time t.

이다.And, Equation 7 is

am.

이다.Also, Equation 8 is

am.

100 : camera
200: information processing device

Claims (4)

a camera for taking an image of a predetermined target; and
an information processing device for determining whether a fire has occurred based on a result captured by the camera; contains,
The camera is
a) The first filter replacement part, which is formed by sequentially forming the filter exchange driving part (A), the transparent part (B), and the infrared cut filter part (C), the transparent part (B), the infrared transmission part (D), and the filter exchange driving part (A) ) formed in this order, the first filter replacement unit and the second filter replacement unit vertically connect each transparent part (B) to the opposite side's transparent part (B) and infrared transmitting part (D). ) or by partially overlapping all of the infrared cut filter part (C),
b) In the case of the weekly monitoring mode, the first filter exchange unit is replaced with an infrared cut filter unit (C), and the second filter exchange unit is replaced with a transparent unit (B) to realize a color image,
c) In the case of night monitoring mode, the first filter exchange unit is replaced with a transparent part (B) and the second filter exchange part is replaced with a transparent part (B) to realize a black and white image,
d) In the case of fire analysis mode, the first filter exchange unit is a transparent part (B), and the second filter exchange part is filter exchanged with an infrared transmitting part (D) to realize a high-temperature analysis image,
The information processing device,
a') when the light corresponding to the first light quantity preset in b) and c) is incident, instructing to replace the filter in d),
b') when the infrared region light incident by a') corresponds to a preset amount of light, determining a fire; Fire detection video surveillance system with a filter, characterized in that. The method according to claim 1,
The information processing device,
c') when the light lower than the second light quantity preset in b) is incident, instructing to replace the filter in c),
The camera is
e) in the case of the first bad weather mode according to the instructions, the first filter exchange unit is replaced with a transparent part (B) and the second filter exchange part is replaced with a transparent part (B) to realize a black-and-white image; Fire detection video surveillance system with a filter, characterized in that. The method according to claim 1,
The information processing device is
c'') In b), compare the sharpness and saturation values of the current color image with the sharpness and saturation values of the last and previous color images of the current color image. and,
The camera is
e') in the case of the second bad weather mode according to the instructions, the first filter exchange unit is replaced with a transparent part (B) and the second filter exchange part is replaced with a transparent part (B) to realize a black-and-white image; Fire detection video surveillance system with a filter, characterized in that. 4. The method according to any one of claims 1 to 3,
The first filter exchange unit and the second filter exchange unit,
one configured at the front end or rear end of the lens unit of the camera; Fire detection video surveillance system with a filter, characterized in that.

Images